David Carlisle writes:
 > > Yes mltex is quite appropriate
 >
 > I wasn't criticising, just checking that I understood you correctly.
 >
 > > for a default format for French as well
 > > as for all european languages and produced glyphs a far better...
 > "better" than what?

let's say all other cm like fonts. Let's compare what is
compareable.

 > Couldn't one produce a VF file that produced
 > identical results to mltex's positioning of the accents?

it should but it is not at this time.

 > I don't think that you'll get universal agreement about

is it really any argument here?

 > >  Since Web2c is a standard, the --mltex option is standard too.
 >
 > while it's reasonable to ask all TeX implementors to implement Knuth
 > specified features like VF support (I could mention some guilty parties
 > at this point:-) I don't think it reasonable to suggest that they
 > "should" implement features just because they happen to be implemented
 > in web2c even if that is by far the most commonly used tex.

nobody is asked to implement what they don't need. Please let
english speaking people, for example, continue to use default cm
if they like it. And let other europeans continue to use the --mltex
option of web2c while they don't need more for their default format.

I don't say that the solution is not going via virtual fonts;
i just say that quality is not sufficient at this time and that
more improvements are necessary: i think to accent placement made
dynamically by the output driver on the basis of \special commands
initiated by linguistic style. No so complicate.

Clearly i refuse any solution which would give me, defaultly,
a less quality output, due to the default (v)fonts in use.

  --bg